为完成大幅面静态图像的采集, 设计了基于多传感器阵列的大视场成像系统。提出了一种改进的基于SIFT算子的图像匹配方法对采集的图像进行匹配, 有效地降低了误匹配率。推导了目标视场角与图像间重叠区域的关系以及成像平台参数和图像间重叠区域的关系, 用于计算图像间的重叠比例, 并只在重叠区域内进行特征点的定位、图像的匹配及拼接, 有效地缩短了图像拼接时间。提出了一种改进的重叠区域渐入渐出融合算法对图像进行融合, 得到了理想的融合效果。

For the acquisition of the large still images,a large field of view imaging system based on multi-sensor array is designed. An improved image matching method based on SIFT operator is used to matching image. The false match rate effectively is reduced. The relationships between overlap region of image and the target FOV and overlap region of imaging and platform parameters are derived, being used to calculate the proportion of the images and position of feature points in the overlap region, so as the image matching and stitching, effectively shorten the time of image stitching. An improved fusion algorithm is used, and the ideal fusion image is obtained.